With the rapid development of mobile communications technologies, Coordinated Multiple Point (CoMP) technology is considered as one of the most promising core technologies in Long Time Evolution-Advanced (LTE-A) technology. Cooperative-Multiple-Input Multiple-Out-put (Co-MIMO), as a typical scenario of the CoMP, has a basic aim of providing services to one or more User Equipments (UEs) through cooperative processing between multiple Base Stations (BSs), so as to effectively eliminate inter-cell interference, and enhance a throughput of cell-edge users and an average cell throughput.
In performing the Co-MIMO, a basic task is determining how to rotationally divide the BSs, so as to form different Cooperation Areas (CAs), and therefore to obtain the best system performances. In view of the task, currently proposed CA division methods substantially include the following several methods, which may be divided into two main types with or without overlaps of the CAs. Typical methods without overlaps of the CAs include, for example, network-centric, network defined and UE-assisted, and typical method with overlaps of the CAs includes, for example, UE-centric.
As for the two types of methods, in the type without overlaps between the CAs, interference influences between the CAs are omitted. Due to the absence of correlation between the CAs, scheduling computation is only performed between the BSs in the CA, and computation tasks between different CAs may be performed in parallel. This solution may effectively eliminate the interference between the BSs in each CA. However, as in this type of methods, mutual interference of edge areas in joint parts of different CAs is not considered, poor user experience is caused in those areas, inter-cell interference exists in large quantity at the same time, and throughput gain that may be obtained by the system is limited.
The type of methods with overlaps of the CAs may effectively avoid the occurrence of the foregoing problem. As overlaps between the CAs are considered in this type of methods, coupling exists between the CAs. Therefore, a scheduling process requires cooperation of all overlapped CAs, and the cooperation is finally extended to the whole network. In a scheduling process of the system, scheduling between all BSs in the system needs to be considered. In the scheduling computation, both intra-CA interference and inter-CA interference are considered, so a good throughput gain may be obtained. However, as in this type of methods, all overlapped CAs are scheduled as a whole, and an ultimate case of such a scheduling is the scheduling of the whole network, the scheduling process is very complex, so that a practical application is very difficult to be achieved.